Technical Product Manager, Applied AI

Clariti Cloud Inc.

1d•CA$130,000 - CA$165,000•Remote

About The Position

Join our mission to provide governments with exceptional experiences so they can do the same for their communities! We empower governments to deliver exceptional citizen experiences. Engineering at Clariti runs on our Agentic SDLC framework: agent specs, reusable prompts, orchestration templates, and eval harnesses that let small pods deliver at multiples of traditional velocity. Until now, that framework served engineering. Your mandate is bigger: treat AI capability as a platform product for the whole company. Think of this as DevOps for AI. DevOps teams do not write the application code; they build the pipelines, guardrails, and golden paths that make every engineer faster. You will do the same for AI: maintain and evolve the Agentic SDLC framework for the engineering and delivery organizations, extend it into our Professional Services practice, and productize the Clariti AI harness, the set of MCP connectors, skills, agent templates, and guardrails that lets non-technical employees in Sales, CX, Finance, People, and PS use AI safely on real work without becoming prompt engineers. You are the second owner of this system, taking over from a founding internal PM who is moving to our flagship product build and who will onboard you and stay engaged through the transition until you are ready to own it fully. The foundations exist. Your job is to scale them from one team's tooling into company infrastructure.

Requirements

5+ years in technical product management, platform engineering, solutions engineering, or applied AI roles, with at least 1 to 2 years shipping LLM or agentic systems into production or into daily internal use.
Hands-on fluency with the current agentic stack: you have personally built with agent frameworks (Claude Code, LangGraph, or equivalents), MCP or comparable tool protocols, and eval-driven iteration. You can read and write an agent spec, not just commission one.
A platform mindset: you measure yourself on other teams' throughput, and you know the difference between building a tool and driving its adoption.
Proven ability to make powerful technology usable by non-technical people, including the judgment for when to build, buy, or simply document.
Strong written communication. Most of your influence will be through specs, playbooks, and enablement content.
Comfort with ambiguity and a small blast radius: this is a pod of few, not a department of many.

Nice To Haves

Experience in regulated or public-sector software, where auditability and defensibility of outputs matter.
Experience in professional services or delivery organizations; you understand utilization, margin, and why consultants distrust tooling that slows them down.
Prior ownership of an internal developer platform, DevEx, or DevOps function

Responsibilities

Own the Agentic SDLC framework roadmap. Prioritize and ship improvements to the agentic SDLC used by product development pods: agent specs, orchestration workflows, reusable prompts, and the eval harnesses that let us trust and improve agent output. Treat evals as the PRD: if we cannot score an agent's output automatically, we cannot scale it.
Extend the Agentic SDLC framework into PS delivery. Partner with the PS leadership team to embed agentic workflows into active implementation projects, from discovery through build. Instrument the before and after so margin impact is measurable, not anecdotal.
Productize the Clariti AI harness for non-technical teams. Ship and maintain the connector layer (MCP integrations into our core systems), a curated skill and template library per function, and the onboarding paths that take an employee from zero to producing real work with AI. Success is measured at the point of use, not in training attendance.
Run the enablement flywheel. Own the AI fluency program end to end: assessments, coaching content, team-level reporting, and the feedback loop from usage data back into the harness roadmap. Enablement is a distribution problem, not a training problem; your job is to put capability inside the workflow where the decision happens.
Own AI vendor and model strategy for internal use. Evaluate models, harnesses, and tools; manage spend; keep switching costs low by favoring open protocols (MCP) and portable assets (prompts, specs, evals) over vendor lock-in.
Own governance for agent output. In govtech, agent-assisted work can end up in front of a planning commission. Define the audit trail, versioning, and approval standards for agent-produced artifacts, and make the safe path the easy path.